Alcoholic fermentation, mostly performed by the yeast species Saccharomyces cerevisiae in food and beverages, leads to the generation of ethanol and carbon dioxide with an outstanding yield of 92% on average. Due to global changes, ethanol content in wine has been looked in-depth in the last two decades but still remains a major research topic. Generally studied in the aim to decrease it, some particular conditions make its increase interesting to design. Because of legal and societal brakes, use of GMO is limited in this context and therefore a possible approach to develop new strains without genetic engineering is the use of adaptive laboratory evolution (ALE), which requires the existence of a potential variation on the phenotype wanted. Thus we detail here a screening strategy to assess variation of production of five primary metabolites in wine-like laboratory conditions using 51 strains from various genetic origins. Our results highlight a low, but still measurable, ethanol production diversity among strains and links between other minor metabolites and genetic origin. These results led us to elaborate an ALE strategy using 2-deoxyglucose, a glucose analogue which acts as glycolysis inhibitor, to impact ethanol production of three wine strains. After 200 generations, our strategy does not reach our objective of ethanol production improvement. However, a remarkable evolution of inhibitor resistance is observed for evolved lineages, from an ancestor with 50% growth reduction to no impact for the best evolved lineage. These results open path for other phenotypic and genetic characterisation of resistance to glycolysis inhibitors